PROJECT SUMMARY - OVERALL Childhood Cancer is the second most common causes of childhood deaths. Cancers in children are highly distinct from those in adults. Causes, mechanisms and therapeutic approaches cannot be extrapolated from the study of adult malignancies. To impact the burden of pediatric cancers requires the specific study of pediatric cancers. We therefore propose the in depth characterization of three specific subtypes of pediatric malignancies which, together, account for over 50% of all pediatric cancer deaths: high grade glioma (pHGG), high risk neuroblastoma (NB), and very high risk acute lymphoblastic B-cell precursor leukemia (VHR-ALL). All three tumor types share a characteristic of typically initial response to therapy, followed by the emergence of resistance and refractory disease. We will map molecular and cellular changes in tumor cells, microenvironment and the immune system using comprehensive multi-dimensional single-cell and in situ technologies associated with two critical transitions: initial response, and emergence of resistant disease ? both high-priority transitions. Treatment modalities will include standard chemotherapy, molecular targeted therapies and chimeric antigen receptor (CAR-) T cell therapy, capturing the cutting edge of current cancer therapy. The final product will impact global childhood cancer research in the following manner: 1) Tumor Atlases: atlases will provide a highly user friendly, publicly available, searchable database of the most comprehensive multi-omic, single cell analysis of the three most lethal childhood cancers. Molecular data will be richly annotated with access to additional pathological evaluation, clinical and outcomes data, and grant access to source data such tumor imaging. 2) Computational methods: in addition to the data, the critical computational tools and pipelines used in this project will be available to the research community. These include methods and pipelines for processing multi-omics and in situ data, inference of cancer phylogeny, inference of cell- and clone-specific pathways, methods for inferring cell-cell crosstalk, as well as database algorithms for the query, exploration and visualization of highly complex data. 3) Access to biospecimens for follow up studies: biospecimens collected in this project will be banked and made available to the biomedical research community. These include tissue sections, viably frozen specimen, and patient derived xenograft (PDX models). In summary, the proposed project will seek to address a major public health need in pediatrics, broadly impact the entire research community, and jumpstart target discovery based on a sophisticated understanding of the key molecular circuits that drive pediatric cancer.